Defined as the analysis of material properties in industrial manufacturing processes, process analysis has been performed for several decades in a wide variety of industries. These industries include chemical, petrochemical, petroleum, pharmaceutical, food & beverage, pulp & paper, and agricultural. A former common implementation of process analysis consisted of manually extracting samples from a process and carrying the samples to a laboratory for analysis. Over time, process analysis evolved from off-line analysis to a continuous on-line analysis where samples are extracted by automated sampling systems and carried in slip streams to process analyzers.
The primary advantage of on-line process analysis is the reduction of the time interval between sample extraction and data generation. The faster response time provides greater control of manufacturing processes leading to increased product yield, improved product quality (consistency), reduced in-process inventory, reduced operating and maintenance workforce, reduced energy consumption, reduced consumption of raw material inputs, and reduced production of waste streams.
Several instruments are currently used for industrial process monitoring. Gas chromatographs (GCs), for example, measure differences in molecular mobility to identify multi-component samples. GCs have high specificity and high sensitivity. They require shielded enclosures for protection from the environment, a supply of column gas, frequent maintenance, and water trapping especially in corrosive applications. These instruments are widely discussed in published literature.
Infrared (IR) instruments rely on material absorption to analyze samples. IR instruments include Fourier Transform Infrared (FTIR) analyzers, IR dispersive analyzers, and non-dispersive IR (NDIR) analyzers. Non-dispersive instruments include filter and non-filter based instruments. IR instruments have displaced other types of instruments due to higher speed, sensitivity, and specificity. IR instruments typically induce a net change in dipole moment in the molecules of a sample as a result of rotational or vibrational motion. The method works well for many species, but fails for homonuclear species such as nitrogen, oxygen, chlorine, hydrogen, and fluorine that cannot have a net change in dipole moment.
Electrochemical sensors provide other means for quantifying species concentrations. These types of sensors are typically limited to the measurement of a single species and often supplement IR methods.
An alternative approach for industrial process monitoring includes the use of Raman methods. Raman spectroscopy is based on the inelastic scattering of light off molecules. As a process analysis technique, Raman spectroscopy has advantages over other techniques as it requires no sample extraction or sample preparation, can perform continuous in-situ quantitative measurements, can analyze pipe content through a sight window, can detect molecules that other techniques cannot, and is unaffected by water molecules.
As a result, Raman spectrometers have found a niche in the market where no other viable solutions exist. Despite these advantages, broad adoption of Raman spectrometers has been hindered because they are very expensive to buy, install and maintain, require frequent calibrations and skilled operators and, in general, lack the robustness necessary to operate in harsh plant environments.
In order for a Raman instrument to be widely accepted for industrial process monitoring, it must have low cost and have high performance. The present invention uses fewer and more readily available components than other Raman instruments, and is easily manufactured and adapted to different applications. It eliminates the use of optical fiber hence achieves high optical throughput. The invention also uses increased amplification with robust multi-stage photon-to-electron amplifiers, and optimized optical filter designs. Further, the invention can withstand tough industrial conditions and uses low cost and wavelength stabilized laser sources.
Raman spectrometers are part of a general class of instruments called optical analyzers. Optical analyzers are generally based on one of six phenomena: absorption, fluorescence, phosphorescence, scattering, emission, and chemoluminescence. These phenomena can occur in the ultraviolet, visible, and infrared portions of the spectrum. A typical instrument contains five basic elements: a radiation source, a sample container, a spectral element to look at a specific region of the spectrum, a detector that converts photons to electrons, and a signal processor. Raman is classified as a second order scattering process in that Raman scattered photons are created from the inelastic interaction of incident light photons with the molecules of the sample. These second order photons are weak, typically 106 to 107 times less intense than first order elastically-scattered photons.
U.S. Pat. Nos. 4,648,714, 4,784,486, 5,521,703, and 5,754,289 use Raman scattering to perform gas analysis. Gases flow through a section of tube while a laser beam is directed into it. These inventions require a slip stream or redirection of the sample away from a pipeline or reactor. Most use a filter wheel in conjunction with a single detector. U.S. Pat. No. 5,521,703 differs slightly from the other three in that its multiple detectors are arranged along the length of gas sampling cell within a laser resonator configuration. U.S. Pat. No. 5,754,289 teaches the use of a filter wheel in conjunction with an integrating sphere for the sample. The related U.S. Pat. Nos. 5,386,295, 5,357,343, and 5,526,121 teach the use of a filter wheel spectrometer coupled to reference and sample elements using fiber optic probes. U.S. Pat. Nos. 5,963,319 and 6,244,753 teach the use of a dispersive spectrometer and fiber optic couplers for industrial process monitoring. Fiber optic couplers are known to limit optical throughput.